**************** 

* Title: Elite Influence on Attitudes about Gender Egalitarianism: Evidence from a Population-Based Survey Experiment
* Journal: Political Behavior
* Authors: Alper T. Bulut and T. Murat Yildirim



*** Set working directory
*cd ""

* Authors' note: The dataset provided for replication includes only the variables utilized in the study. We are unable to share the full dataset
* that includes all other survey questions due to copyrights. The full dataset can be obtained from KONDA Research and Consultancy (www.konda.com.tr)


log using "Elite Influence on Gender Egalitarianism--Replication.smcl", replace



* Figure 1
gen equalofopportunity_abscorrect = (equalofopportunity==5)

***
graph hbar equalofopportunity_abscorrect if AKP==1, scheme(s1mono) over(treatment, relabel(1 "Control" 2 "Treatment")) name(AKP, replace)
graph hbar equalofopportunity_abscorrect if CHP==1, scheme(s1mono) over(treatment, relabel(1 "Control" 2 "Treatment")) name(CHP, replace)
graph hbar equalofopportunity_abscorrect if fem==1, scheme(s1mono) over(treatment, relabel(1 "Control" 2 "Treatment")) name(female, replace)
graph hbar equalofopportunity_abscorrect if college==1, scheme(s1mono) over(treatment, relabel(1 "Control" 2 "Treatment")) name(college, replace)

graph combine AKP CHP college female, scheme(s1mono) xcommon cols(2)



* Figure 2
gen womenpolitics_abswrong = (womenpolitics==1)

***
graph hbar womenpolitics_abswrong if AKP==1, scheme(s1mono) over(treatment, relabel(1 "Control" 2 "Treatment")) name(AKP_2, replace)
graph hbar womenpolitics_abswrong if CHP==1, scheme(s1mono) over(treatment, relabel(1 "Control" 2 "Treatment")) name(CHP_2, replace)
graph hbar womenpolitics_abswrong if fem==1, scheme(s1mono) over(treatment, relabel(1 "Control" 2 "Treatment")) name(female_2, replace)
graph hbar womenpolitics_abswrong if college==1, scheme(s1mono) over(treatment, relabel(1 "Control" 2 "Treatment")) name(college_2, replace)

graph combine AKP_2 CHP_2 college_2 female_2, scheme(s1mono) xcommon cols(2)






****** Table 2
ologit equalofopportunity treatment

ologit equalofopportunity treatment if AKP==1

ologit equalofopportunity treatment if CHP==1

ologit equalofopportunity treatment fem religiosity eduord age age2 urbanrural income econ_wellbeing  

ologit equalofopportunity treatment fem religiosity eduord age age2 urbanrural income econ_wellbeing if AKP==1

ologit equalofopportunity treatment fem religiosity eduord age age2 urbanrural income econ_wellbeing if CHP==1


********* Table 3
ologit womenpolitics treatment

ologit womenpolitics treatment if AKP==1

ologit womenpolitics treatment if CHP==1

ologit womenpolitics treatment fem religiosity eduord age age2 urbanrural income econ_wellbeing  

ologit womenpolitics treatment fem religiosity eduord age age2 urbanrural income econ_wellbeing  if AKP==1

ologit womenpolitics treatment fem religiosity eduord age age2 urbanrural income econ_wellbeing  if CHP==1








* Substantive Impact using Clarify
************ Equality of Opportunity (First DV)
estsimp ologit equalofopportunity fem religiosity eduord age age2 urbanrural income econ_wellbeing treatment 
setx mean 
simqi, pr
simqi, fd(pr) changex(treatment 0 1) level(90)
* Drop the simulated parameters to use the "estsimp" for the second DV
drop b1-b13


estsimp ologit equalofopportunity fem religiosity eduord age age2 urbanrural income econ_wellbeing treatment if CHP==1
setx mean 
simqi, pr
simqi, fd(pr) changex(treatment 0 1) level(90)
drop b1-b13


* Divide the values of "First Difference" for each answer (i.e., absolutely wrong, wrong, etc.) by the baseline probability of the
* respective answer to calculate the percentage change. For instance, among the CHP voters, the treatment condition reduces the
* likelihood of saying "absolutely correct" for the first statement (i.e., "equality of opportunity") by 12.7 percent (-.0817387/.6471322)
* Note that each time we use the "estsimp" command we will get slightly different results as these results are based on simulated probabilities.

estsimp ologit equalofopportunity fem religiosity eduord age age2 urbanrural income econ_wellbeing treatment if AKP==1
setx mean 
simqi, pr
simqi, fd(pr) changex(treatment 0 1) level(90)
drop b1-b13


************ Women's Presence in Politics (Second DV)
estsimp ologit womenpolitics fem religiosity eduord age age2 urbanrural income econ_wellbeing treatment 
setx mean 
simqi, pr
simqi, fd(pr) changex(treatment 0 1) level(90)
drop b1-b13


estsimp ologit womenpolitics fem religiosity eduord age age2 urbanrural income econ_wellbeing treatment if CHP==1
setx mean 
simqi, pr
simqi, fd(pr) changex(treatment 0 1) level(90)
drop b1-b13


estsimp ologit womenpolitics fem religiosity eduord age age2 urbanrural income econ_wellbeing treatment if AKP==1
setx mean 
simqi, pr
simqi, fd(pr) changex(treatment 0 1) level(90)
drop b1-b13




log close
